Goals: This R01 will deploy a smartphone based intervention for bipolar disorder to 1) reduce relapse and symptoms, 2) investigate targets mediating clinical outcomes, and 3) create a longitudinal feature-rich database to better understand illness trajectories and improve treatments. Background: Bipolar disorder is a severe and chronic mental illness. Combined treatment with medications and psychotherapy improves outcomes, but access to therapy is limited. Current therapies do not capture behavioral and other data that could improve our understanding of the disorder and its treatment. They also do not provide real-time feedback which could improve patient self-management and facilitate provider intervention. Finally, limited data exists specifying what behavioral targets mediate psychotherapy benefits. Significance: This study will use a smartphone to increase access to proven strategies that enhance patient self-management and provider intervention. Simultaneously, self-report, behavioral, and application usage data will be collected along with clinical status assessments creating a database for understanding intervention effects and elucidating phenotypes. Finally, changes in proposed targets mediating clinical outcomes will be specifically assessed with validated measures to understand intervention mechanisms. Innovation: 1) A theory-based approach organizes the intervention design and provides tools for investigating intervention mechanisms effecting change, 2) Use of a coach guided by the supportive accountability model will improve adherence and outcomes, 3) Highly personalized wellness plans and tailored feedback reactive to patient status will also improve outcomes, 4) The feature-rich data collected will facilitate understanding how the intervention works and elucidate bipolar disorder phenotypes leading to future improvements in treatment. Preliminary studies: Iterative user-centered design and usability testing elicited feedback on all aspects of the intervention from patients and psychiatrists. Daily self-monitoring and behavioral data collection was successfully tested in a field trial. All aspects of smartphone application design, content, and software are fully operational. Coach training and delivery materials as well as validated assessment tools for clinical outcomes, proposed targets and their determinants are fully developed. The complete intervention is being field tested. Design: A single blind, randomized controlled trial (N = 200, 2:3 ratio of usual care vs usual care and intervention) will be conducted in clinical practices around Chicago and Minneapolis. Population: Adult bipolar 1 disorder patients not experiencing a current mood episode with a history of two acute episodes in the last two years who are in treatment with a psychiatrist. Outcomes: The primary clinical outcomes will be time to relapse, and the secondary outcomes will be percent time symptomatic, symptom severity, and quality of life. Target outcomes will be medication adherence, sleep duration, routine regularity, and management of early warning signs.